Metraipolis › archive › dc_574fef12a79e

Kinematic Causal Attribution: Black-Box Explanations of Motion versus Geometry for 4D Point-Cloud Video Models — arXiv:2610.39147

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.39147
License (as archived)
CC0-1.0
Archived
2026-11-01T02:41:49.776Z
SHA-256 of the archived text
9406142452b14f7be32a28463cf4240ad3c117e34f5c8368179c4ecf71a3866a
Stable URL of this page
https://metraipolis.com/documento/dc_574fef12a79e

How to cite

Kinematic Causal Attribution: Black-Box Explanations of Motion versus Geometry for 4D Point-Cloud Video Models — arXiv:2610.39147 — arXiv (en), https://arxiv.org/abs/2610.39147 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_574fef12a79e (Metraipolis archive dc_574fef12a79e, sha256 9406142452b14f7be32a28463cf4240ad3c117e34f5c8368179c4ecf71a3866a)
BibTeX
@misc{dc574fef12a79e,
  title = {Kinematic Causal Attribution: Black-Box Explanations of Motion versus Geometry for 4D Point-Cloud Video Models — arXiv:2610.39147},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.39147},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_574fef12a79e; sha256 9406142452b14f7be32a28463cf4240ad3c117e34f5c8368179c4ecf71a3866a},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.39147. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

4D perception models, which classify 3D point-cloud videos, are being built into robots, assistive monitoring and AR, yet before deploying one a practitioner cannot check which cue it relies on: how a body part moves, or what shape it has. 3D attribution is single-frame, video attribution is confined to RGB pixels, gradient maps need white-box access, attention is absent from state-space backbones, and point-level perturbations assume a point correspondence across frames that real depth and LiDAR data lack. We present Kinematic Causal Attribution (KCA), a black-box, interventional attribution method that queries only the model's outputs. Its freeze operator removes a region's motion while keeping its shape, and a shape-free proxy removes the shape while keeping the motion. Together they span a motion$\times$geometry cube for every tracked region, whose order-symmetric Shapley decomposition is complete and whose interaction is measured rather than assumed away. The operators rewrite per-frame point sets, so they need no correspondences, and a context-averaged estimator handles structures that span several regions. We validate KCA on the official PSTNet, P4Transformer and Mamba4D trained on MSR-Action3D. The decisive test is an audit: two binary tasks are built from real clips whose classes differ only in the motion or only in the shape of the same arm, and the motion share of the arm's importance follows the defining cue on all three backbones (0.83 vs. 0.60 on P4Transformer, 0.86 vs. 0.30 on PSTNet, 0.92 vs. 0.61 on Mamba4D), a distinction no single attribution map can make. The split costs no faithfulness: the necessity maps have the lowest deletion AUC on P4Transformer and PSTNet, tie with the best baseline on Mamba4D, and lose their correlation with the trained model's once the weights are randomised. Skeleton ground truth, re-sampling stability and cost complete the validation.

Tema degli esploratori: statistica e scienza dei dati (elenco arXiv cs.CV, nuovi lavori del giorno).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2610.39147 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-11-01T02:16:43.907Z dalla pagina https://arxiv.org/list/cs.CV/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).